(213 days)
SwiftMR is a stand-alone software solution intended to be used for acceptance, enhancement and transfer of all body parts MR images in DICOM format. It can be used for noise reduction and increasing image sharpness for MR images. SwiftMR is not intended for use on mobile devices.
SwiftMR, is software used as a Medical Device (SaMD) consisting of a software algorithm that enhances images taken by MRI scanners. The device only processes DICOM images for the end User and is intended to be used by radiology technologists in an imaging center, clinic, or hospital. The device's inputs are MRI images in DICOM format. The deep learning algorithm produces enhanced images as outputs with reduced noise and increased sharpness in DICOM format. The deep learning algorithm performs noise reduction with the ability of adjusting the denoising level 0 to level 8, and sharpening filter performs the sharpening function with the ability of adjusting the sharpness level from level 0 to level 5. SwiftMR provides an automatic image quality enhancement function for MR images acquired in various environments. SwiftMR can only be used for professional purposes and is not intended for use on mobile devices.
Here is the requested information about the acceptance criteria and study proving the device meets them:
Acceptance Criteria and Device Performance
| Metric | Acceptance Criteria | Reported Device Performance |
|---|---|---|
| Noise Reduction | Signal-to-noise ratio (SNR) of SwiftMR-processed images is increased by 40% or more for at least 90% of the dataset for level 1, with an incremental 1% increase per each level. | The test passed, indicating the device met or exceeded this criterion. |
| Sharpness Increase | Full Width at Half Maximum (FWHM) of a selected region of interest (ROI) is decreased by: |
- 0.13% (deep learning model)
- 0.43% (filter level 1)
- 1.7% (filter level 2)
- 2.3% (filter level 3)
- 3.6% (filter level 4)
- 4.5% (filter level 5)
for at least 90% of the dataset. | The test passed, indicating the device met or exceeded these criteria. |
Study Details
-
Sample size used for the test set and the data provenance:
- Sample Size: Not explicitly stated as a single number for the entire test set. However, the validation dataset included retrospective clinical images covering a wide range of conditions.
- Data Provenance: Retrospective clinical images from various manufacturers (SIEMENS, GE, PHILIPS, CANON, ESAOTE, FONAR, FUJIFILM), field strengths (0.25T, 0.6T, 1.5T, 3.0T), anatomical regions (Body, Cardiac, Neuro, Musculoskeletal), and protocols (T1, T2, T2*, FLAIR, PD, DWI, MRA). Demographics included adults (22-93 yrs, 88.4%) and pediatrics (0-21 yrs, 11.6%), with 44.8% male and 55.2% female. The dataset also included images with up to 50% time reduction for reduced scan time images. Importantly, the validation dataset included data from sources not included in the training dataset to demonstrate performance is not hindered by site variability.
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts: Not specified in the provided document. The document describes quantitative metrics (SNR increase, FWHM decrease) as acceptance criteria, suggesting a technical ground truth rather than expert interpretation of a specific condition.
-
Adjudication method for the test set: Not applicable based on the description. The acceptance criteria are based on objective, quantitative measurements (SNR and FWHM changes) derived from the image processing algorithm itself, rather than subjective expert consensus.
-
If a multi-reader multi-case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance: An MRMC comparative effectiveness study is not mentioned as part of the validation for this 510(k) submission. The validation focuses on the standalone performance of the algorithm in enhancing image quality based on objective metrics.
-
If a standalone (i.e., algorithm only without human-in-the-loop performance) was done: Yes, the performance tests described (noise reduction and sharpness increase) assess the standalone performance of the SwiftMR algorithm. The device "only processes DICOM images for the end User" and "performs image processing in the background automatically."
-
The type of ground truth used (expert consensus, pathology, outcomes data, etc.): The ground truth for the performance validation appears to be technical/quantitative metrics (SNR and FWHM measurements) derived from the images themselves, rather than clinical outcomes, pathology, or direct expert consensus on diagnostic accuracy. The aim is to demonstrate that the device effectively reduces noise and increases sharpness as measured objectively.
-
The sample size for the training set: Not specified in the provided document.
-
How the ground truth for the training set was established: Not specified in the provided document. The algorithm uses a deep learning model, and its parameters were "obtained through an image guided optimization process," but the specifics of the training data's ground truth (e.g., if it involved artificially generated noise, paired noisy/clean images, or expert-labeled image quality) are not detailed.
§ 892.2050 Medical image management and processing system.
(a)
Identification. A medical image management and processing system is a device that provides one or more capabilities relating to the review and digital processing of medical images for the purposes of interpretation by a trained practitioner of disease detection, diagnosis, or patient management. The software components may provide advanced or complex image processing functions for image manipulation, enhancement, or quantification that are intended for use in the interpretation and analysis of medical images. Advanced image manipulation functions may include image segmentation, multimodality image registration, or 3D visualization. Complex quantitative functions may include semi-automated measurements or time-series measurements.(b)
Classification. Class II (special controls; voluntary standards—Digital Imaging and Communications in Medicine (DICOM) Std., Joint Photographic Experts Group (JPEG) Std., Society of Motion Picture and Television Engineers (SMPTE) Test Pattern).